RIGS-TO-REEFS AND PORPOISES: PRÉCIS OF EVIDENCE TO SUPPORT LEAVING OFFSHORE INSTALLATIONS IN SITU AT THE END OF THEIR OPERATIONAL LIFETIMES V. L. G. Todd1,2, Jane C. Gardiner1, & I. B. Todd1,2

1Ocean Science Consulting Ltd., UK, 2Institute of Sound and Vibration Research (ISVR), Southampton University, UK

ABSTRACT

Condensed highlights from a decade of Marine Mammal Observer (MMO) and Passive Acoustic Monitoring (PAM) projects carried out from stationary and on-tow Oil & Gas (O&G) exploration jack-up drilling-rigs and production platforms is presented. The majority of work was undertaken as routine monitoring and mitigation permit requirements, but a partial aim was to investigate the hypothesis that, because of a potential rig-as-a-reef effect, echolocation activity of harbour porpoises (Phocoena phocoena) is likely to be higher at offshore installations compared with the open sea. Porpoise activity was monitored visually and acoustically using MMOs, real-time PAM, and a brief, controlled experiment using T-PODs – autonomous, static, echolocation-click detectors. Throughout the last decade, porpoises were detected visually, acoustically, and consistently at varying levels, at all offshore locations undergoing routine operational activities. The controlled T-POD experiment gave reasonable evidence to support the hypothesis that porpoises may target offshore installations compared to the open sea, but further long-term, replicated and controlled visual and acoustic experiments using recently-optimised mooring techniques are required to ‘put the theory to bed'. In conclusion, this paper puts forward the notion that decommissioning of O&G installations in the North Sea may remove valuable porpoise foraging habitat and thus may have implications for the long-term survival of this listed and threatened species.

NOMENCLATURE ANOVA Analysis of Variance CC Correlation Coefficient C-POD Digital porpoise click detector CTD Conductivity, Temperature, Depth

[sensor] dB Decibel DPM Detection Positive Minutes GIS Geographical Information System HMI Head of Mining Installation Hz Hertz JNCC Joint Nature Conservation

Committee kHz Kilohertz m Metre MMO Marine Mammal Observer n Number of samples NBHF Narrow Band High Frequency O&G Oil & Gas P Probability PAM Passive Acoustic Monitoring PPM Porpoise Positive Minutes s Second [time] SD Standard Deviation SL Source Level

T Sum of Ranks (Mann-Whitney) T-POD Analogue porpoise click detector μPa Micro Pascal V Volt VSP Vertical Seismic Profile .wav Wave file 1. INTRODUCTION

Cetaceans (whales, dolphins and porpoises) use different sound frequency bands for communication, foraging, navigation, threat detection/avoidance and a range of activities within the wider social group such as cohesive actions, warnings and maternal relationships [1]. In most cases, the hearing range of marine mammals is less well understood, but it is assumed generally that, with the exception of some porpoise species, animals hear over similar frequency ranges to the sounds that they produce. The reduction of anthropogenic (man-made) noise levels is thus a focus of both industry and regulators, and accordingly, considerable research is being undertaken currently in this field.

Compared to other Oil and Gas (O&G) activities such as seismic

exploration/profiling and conductor driving, drilling and production noise emanating from fixed, metal-legged offshore installations is considered generally to be relatively low level (120 dB re 1 μPa) and in the frequency range of dominant tonals (0.02–1.4 kHz, with little radiated noise over 8 kHz [2].

Odontocetes (toothed cetaceans) vocalise and detect mainly mid–high frequency sound (see chapters 1 and 8 in Todd et al. [1] for a review of hearing and vocalisation ranges in odontocetes). With respect to offshore drilling and production, there are various short and long-term operations that may be detected by, and/or influence the behaviour of odontocetes, and others that are less likely to elicit responses. For example, in the short term (hours c.f. months), supply boat, conductor driving, and Vertical Seismic Profiling (VSP) operations have the potential to produce frequencies in the range of odontocete hearing. Like pile driving during windfarm installation, conductor driving in particular has the potential to cause physical damage, masking of important sounds and/or habitat exclusion effects; however, longer-term (2–3 months), non-impulsive, low Source Levels (SLs) and frequencies produced during drilling and production operations are unlikely to cause physical damage, masking of important sounds and/or habitat exclusion effects of odontocetes [1]. Mysticetes (baleen whales) utilise sound in lower frequency ranges, and may be affected differently.

In the North Sea, the harbour porpoise (Cetacea: Phocoena phocoena) is threatened by starvation, climate change, unsustainable incidental catches in fishing gear, pollution and potential exclusion from foraging habitat by windfarm pile-driving activities [3-6]; however, it has long been established that offshore O&G installations and their associated 500 m fishing exclusion zones, act as artificial reefs and protected habitats for fish, coral and other species assemblages [7-9]. Porpoises also forage within 300 m of installations, predominantly at night [10].

This paper presents condensed highlights from a decade of porpoise visual and acoustic activity around offshore O&G installations,

both in situ and on-tow; combined megafaunal species datasets are presented elsewhere [11]. 2. METHODS

In the North Sea, between 2004 and 2014, annually and throughout all seasons, visual (Marine Mammal Observer; MMO) and acoustic (Passive Acoustic Monitoring; PAM) surveys of marine mammals were undertaken from various offshore O&G jack-up exploration drilling rigs and production platforms, whilst stationary and on-tow. Production platforms had been in situ from 4–16 years. Jack-up drilling rigs either drilled wells at novel locations, or drilled additional wells whilst attached to platforms. Locations of the two stationary production platforms (PA and PB) are shown in Figure 1, along with six monitoring locations (L1 – L6) and five tow routes (1–5). ‘Locations’ refer to areas where monitoring was undertaken from stationary jack-up drilling rigs, pre-, post or independent of tows, and ‘tows’, when monitoring was carried out from a mobile jack-up in transit from one position to another.

Figure 1. Platform (P) and monitoring locations (L), and tow routes (lines). Arrows indicate direction of rig movement.

Full installation technical specifications are omitted for brevity, but typical three and

four-legged jack-ups, and small production platforms are listed in [10]. Real-time towed PAM results are presented, but considerable datasets generated through annual autonomous, underwater click detector T/C-POD (Chelonia Ltd.) observations are omitted, with the exception of one short, controlled T-POD experiment.

Underwater noise measurements were undertaken, but results are presented elsewhere [2,12]. Modelled and empirical oceanographical and weather (e.g. Conductivity, Temperature, Depth – CTD – significant wave height, wind speed, etc.) and operational logs were collected by MMOs and PAM Operators for all monitoring periods, but again for brevity, only forecast significant wave height and wind speed are presented, along with Beaufort sea state, recorded by field personnel.

2.1 MMO

Marine Mammal Observers used hand-held 7x50 binoculars or the naked eye to scan the horizon from 0° to 90° on either side of the installations’ heading; viewer location varied according to installation type and barge height (15–40 m). Joint Nature Conservation Committee (JNCC)-trained MMOs conducted visual observations throughout daylight hours using standard JNCC techniques [13]. Industry-standard range (www.osc.co.uk/tools/range-estimation-using-reticle-binoculars) and bearing (www.osc.co.uk/tools/bearing-estimation) estimations for sightings were employed. Details of all sightings including time, location, range, species, number of individuals and behaviour, were entered into JNCC ‘Record of Sightings’ forms. No visual observations were undertaken in weather conditions exceeding a Beaufort sea state 6.

2.2 REAL-TIME PAM

Towed, real time PAM was deployed on two tows and from two production platform-jack-up drilling rig complexes when stationary. A 200 m potted hydrophone array (Seiche Measurements Ltd) was used to monitor for porpoises. The array consisted of four transducers arranged in three pairs; spacing

between hydrophones one and two was 0.25 m, and remaining pairs were spaced 1.2 m apart. Each spherical ceramic transducer was connected to a 40 dB preamplifier with band pass filter. Hydrophones one to three were broadband, with band pass filter -3 dB points at 2 kHz and 200 kHz. Hydrophone four was low frequency with band pass filter -3 dB points at 75 Hz and 30 kHz. Hydrophone sensitivity was -166 dB re 1V/uPa and -157 dB re 1V/uPa across the relevant pass bands for the broadband and low frequency elements respectively. Frequency response was flat across the pass band for both. A depth sensor was located at the end of the array. PAMGuard (version 1.12.05) was used to process files.

2.3 T-POD

In January 2010, at a carefully selected control (open sea) location 3 km away from a jack up rig situated at L4, a 4 tonne special marker buoy was moored on the sea bed with a 4.6 tonne block of concrete. Over a 39-day period, using the exact same instruments, version types, methods, and settings given in [10], one version 3 T-POD (ID 408) was attached to the control (open sea) mooring, and one version 3 T-POD (ID 406) from over the side of the jack-up at water depth of 25 m. Acoustic levels of porpoise activity were compared between the jack-up rig and control (open sea) to test the hypothesis that, because of the attraction of potential prey species (rigs-to-reefs), porpoises may prefer the jack-up to the open sea. Restricted budget and time prevented the use of further controls and replicates.

2.4 DATA ANALYSIS

2.4 (a) MMO

Porpoise sightings and MMO effort data are presented; no seasonal comparative statistical analysis has been undertaken, and sightings are not compared quantitatively to weather/sea state conditions.

2.4 (b) Real-time PAM

High frequency sounds were sampled using National Instruments 6251 sound card,

and a sampling rate of 500 kHz. Porpoise echolocation data were processed using the click detector module in PAMGuard, with settings optimised for porpoises (pre-filter: high-pass Butterworth filter set to 40 kHz; trigger filter: Butterworth band-pass filter set to 125 – 150 kHz, trigger threshold 10 dB).

PAMGuard recorded continuously at all times; data were recorded as .wav files. All recordings were viewed retrospectively in PAMGuard’s viewer, by two independent PAM Operators. All clicks identified as a porpoise by PAMGuard were examined to ensure correct classification. Clicks were defined as porpoise if they satisfied the following criteria: 1) significant energy around 130 kHz band [14-16]; 2) polycyclic waveform resembles that of published data for porpoises [e.g. 14,15,17]; and 3) and narrow band with a -3dB bandwidth of 6–26 kHz [14-16]. A click event was defined as two or more clicks separated by <60 s. Number of minutes in which porpoise clicks were detected (Porpoise Positive Minutes, PPM) were also calculated, and locations of detections on tow were plotted in Geographical Information Software (QGIS 2.8.1).

2.4 (c) T-POD

T-POD software version used was the same as [10] for consistency. Only CetHi trains were analysed, which is the designation TPOD.exe uses for trains most likely to have been produced by the target species (porpoises). Data for each porpoise train were exported from T-POD.exe into Microsoft ExcelTM for analysis. Detection Positive Minutes per day (DPM/day) was used as an indication of porpoise activity at the jack-up rig and the open sea control. A minute was classified as porpoise positive if it contains any identified porpoise train(s), and a day is a standard 24-hour (diel) period.

Statistical tests were performed using SigmaStat v.3.1 (Systat software Inc., CA, USA). All train dataset were non-normally distributed (Kolmogorov–Smirnov tests, P<, 0.05), and logarithmic and arcsine transformations failed to normalise data. Non-parametric Mann-Whitney Rank Sum tests and Kruskal–Wallis, one-way Analysis of Variance (ANOVA), with the appropriate post hoc tests,

were therefore employed to assess significant differences for the indicators of porpoise activity between the jack-up rig and control (open sea).

3. RESULTS

100% of both real-time and fixed PAM gear were recovered over the decade (i.e. no losses).

3.1 MMO

Marine Mammal Observer effort and forecast weather for stationary installations and rig tows is given in Figure 2 and Figure 3 respectively. Figure 4 displays MMO sightings of porpoises from stationary and on-tow installations, over ten years.

A total of ca. 226 hours of visual observations were carried out by MMOs on 44 days over ten years, in all seasons. Effort varied considerably, restricted mostly to short periods before operational start-up periods (e.g. drilling, conductor driving, VSP, etc.), as per permit requirements.

In summary, over a decade, a total of 26 harbour porpoise sightings, of 49 individuals were recorded at the various locations; 3 from the sides of stationary installations (Figure 2) and 46 from rigs on-tow (Figure 3). Range to animals varied from 100–2,012 m

3.1 (a) OBSERVATIONS FROM STATIONARY INSTALLATIONS

Position of all fixed installation sightings is shown in Figure 4. There were only two porpoise sightings, of three individuals, from stationary installations over ten years. Both occurred incidentally i.e. MMOs were not conducting dedicated watches.

In 2005, a mother and calf were sighted incidentally from PA at a range of 100 m by the Head of Mining Installation (HMI), and a clear photograph was taken (not presented here). The platform was engaged in routine production activities. Weather conditions were not recorded, but from the photo, it is estimated to be a Beaufort sea state 2.

In 2014, a second incidental sighting of a porpoise occurred from the side of PA at a range of 100 m. The platform was engaged in routine production activities. Sea state was not recorded at the time of the sighting, but mean forecast weather showed significant wave height was < 0.5 m and wind speed was 4 m/s, which equates roughly to a Beaufort sea state 3.

Figure 2. Forecast weather, MMO effort and porpoise sightings recorded from static installations, including both jack-up drilling rigs, and rig-platform complexes.

3.1 (b) Observations on-tow

Visual observations were conducted during five tows. Sightings of porpoises were made during tows in 2004, 2012 and 2014, shown in Figure 3 and Figure 4.

In 2004, a single dedicated sighting of a porpoise occurred at a range of 350 m from the side of a jack-up rig situated at L2 (Figure 4). The rig had just completed a tow, and was engaged in running of the 30” conductor. Weather conditions were excellent, with a Beaufort sea state of 1 recorded at the time of the sighting.

In 2012, the jack-up rig was under tow from L5 to PA (Figure 1). On a single day (15/01/2012), 22 dedicated sightings of 44 harbour porpoises were recorded. The first sighting occurred 102.92 km from L5 and the last sighting 33.14 km from PA, where the jack-up was scheduled to drill a new well; sightings occurred over a distance of 49.96 km. Mean porpoise group size was two, ranging from one to five animals (SD 1.203). It is worth noting that this particular rig tow occurred during the day in Beaufort sea state 0-2 (ideal porpoise sighting conditions) (Figure 3).

In January 2014, a single harbour porpoise was sighted on a dedicated watch from a jack-up rig, towards the beginning of a tow from PB to PA (Figure 1). The sighting occurred 7.5 km from the starting location. Sea state was a Beaufort 2 at the time of the sighting, but degraded to a Beaufort 4 thereafter.

Figure 3. Forecast weather, MMO effort and porpoise sightings recorded whilst on tow.

Figure 4. Location of static and on-tow visual harbour porpoise sightings.

3.2 REAL-TIME PAM Real-time PAM was collected during two

rig tows in 2014. A total of ca. 160 hours of PAM recordings were made, 63 hours in January, and 97 hours in June. Hours of recording during daylight and darkness (as specified by sunrise and sunset times), number of harbour porpoise clicks detected, and operational activity at the time of detections is summarised in Figure 5.

Figure 5. Daytime (D) and night time (N) PAM effort, number of harbour porpoise clicks recorded, and operational activity at time of detection.

A total of 308 individual clicks were identified as porpoises. Divided into monitoring period, 13 click events or 29 PPM were recorded in January, and 7 click events or 20 PPM in June.

All detections in January were recorded from a stationary jack-up drilling rig/production platform complex situated at PB, prior to a rig tow. Six out of the seven click events recorded in June occurred also from a jack-up drilling rig whilst joined to PA, but a single click event (128 clicks; 9 PPM) occurred towards the end of a tow (PA-PB), 1.2 km from PB, but prior to any platform joining activity. Detection start and end location is shown in Figure 6.

Figure 6. Start (s) and end location of the on-tow PAM detection.

In January, two single clicks and one click event occurred during daylight hours, which equated to 5 PPM; sea state was Beaufort 5, and no visual observations were carried out. In June, 57% of the click events (17 PPM) occurred during daylight hours; MMOs were on watch for two of the click events, but no animals were sighted visually. Sea state was recorded by the MMO’s as a Beaufort 3 during the acoustic detection on 28/06/14 and a Beaufort 5 on 30/06/14.

3.3 T-POD Figure 7 shows the DPM/day for both the

T-PODs at the jack-up rig compared with the open sea control. There were significantly higher numbers of porpoise acoustic detections (DPM/day) at the jack-up rig compared to the open sea control (Mann-Whitney Rank Sum Test T = 1383.000, P = 0.023, control: n1 = 40,

median = 1.000, 25 % = 0.000, 75 % = 5.000; rig: n2 =40, median = 4.000, 25 % = 1.000, 75 % = 8.000). The pattern of porpoise detections was not similar at the two locations (Spearman Rank Order Correlation CC = -0.0177, P = 0.912).

0

2

4

6

8

10

12

14

16

18

20

21/0

1/20

1023

/01/

2010

25/0

1/20

1027

/01/

2010

29/0

1/20

1031

/01/

2010

02/0

2/20

1004

/02/

2010

06/0

2/20

1008

/02/

2010

10/0

2/20

1012

/02/

2010

14/0

2/20

1016

/02/

2010

18/0

2/20

1020

/02/

2010

22/0

2/20

1024

/02/

2010

26/0

2/20

1028

/02/

2010

DPM

/day

Date

Control T-POD 408Rig T-POD 406

Figure 7. Comparison of jack-up rig and open sea control, for T-POD porpoise Detection Positive Minutes per day (DPM/day). 4. DISCUSSION

Visual and acoustic detections reported within this paper across ten years, from jack-up drilling rigs and production platforms show that porpoises are present in the near vicinity of exploration drilling rigs and production platforms during routine operations.

Porpoises were detected visually in relatively low numbers considering observer effort. Given that porpoises forage actively around and between the legs of installations within 300 m, and with high encounter rates per hour [10], more porpoises were likely to have been present around installations than were actually seen or detected. The lack of visual detections was likely twofold. Firstly, porpoises around O&G installations are more active at night [10], and secondly, results presented here corroborate other studies [18,19] that show visual detections are significantly less likely in Beaufort sea state > 2. Weather offshore is clement rarely, and operations continue unhindered in conditions far in excess of this.

Similarly, with regards to PAM, the extensive T/C-POD datasets (a fraction of which are presented here) suggest that more porpoises are present, than were detected around installations. Firstly, porpoises’ Narrow Band High Frequency (NBHF) echolocation

clicks attenuate rapidly, so vocalising individuals need to be close to the hydrophone (<300 m) to be detected. Directionality is important also, as clicks from animals facing away from hydrophones will either be missed entirely, or only partially recorded. Equally, characteristics of off-axis clicks (SL, frequency, etc.) differ from those recorded on-axis, and maybe unrecognisable as porpoises by PAMGuard. In addition, lateral and vertical head movements of echolocating porpoises [20,21], mean it is likely that click trains were only recorded partially, which may explain why number of clicks per click event ranged from 2–128 (mean 14, SD 27).

Masking of received biosignals by localised operational noise also reduces PAM efficacy. This can be minimised by optimising deployment location, but in an industrial setting, location and length of hydrophone cable deployment is restricted, and avoidance of operational noise unavoidable. Moreover, un-streamlined rigs are not designed specifically for towing, which creates wake and turbulence around hydrophones, resulting in flow-noise, and oscillation in transducer depth, further reducing detection probabilities.

This paper presents the first evidence of a mother and calf around a platform (in this case, ca. 250 km from land) and is further proxy evidence that porpoises are using O&G installations to feed. This is because porpoises are small, with limited body fat, and relatively high metabolic rates compared to other cetaceans. They have a large surface area over which to lose heat to the environment, where seasonal water temperatures in this study ranged from 4–16ºC. Porpoises can survive on lipids in their blubber as their sole source of nourishment for about a week without feeding (H. Koopman pers. comm.); a lactating female even less. There is therefore considerable pressure on these animals (especially females) to feed at rates higher than larger cetaceans with lower metabolic rates. Food consumption in porpoises varies between 4–9.5% [22] and 7.5–10% body weight daily [23], representing between 8,000 and 25,000 kJ/day [22]. The small size of porpoises does not enable them to carry large energy stores [24], so their patterns

of movement are likely related strongly to distribution of their prey. It is thus highly probable that a female and calf so far away from shore, has targeted the 16-year-old platform as a known and temporally-spatially predictable feeding station with which to provide a reliable milk source for her calf.

When rigs left/reached destinations (or were on-tow), there is a possibility that porpoises detected visually and acoustically were following jack-up rigs as mobile feeding stations; however, without individual recognition, perhaps in the form of porpoise satellite tagging experiments, this hypothesis cannot be tested. It is entirely feasible, given the known sustained swimming speeds of porpoises, ca. 0.7–2.2 m sec-1 [25,26] and typical rig tow speeds of 4–5 kts (2–3 m sec-1) that animals could keep up with rigs on-tow. Further research would need to be undertaken on potential porpoise prey swimming speeds, or known behaviour of rig-aggregated fish species following rig departure from a typical 2–3 month drilling campaign. It is also possible that once a rig has departed a location, porpoise-fish-prey species, particularly those less mobile, such as gobies (Gobiidae) and small gadoids (cod-like species) deprived of rig refugia may be more easily accessible targets for any porpoises remaining in the vicinity. Porpoises detected when rigs were attached to platforms pre rig-tow, or on rig approach to platforms, were likely already in the vicinity, maximising prey availability from the rigs-to-reefs effect from platforms that had been in situ for longer periods.

This paper presents the first real-time acoustic detections obtained from a North Sea jack-up drilling rig whilst using a typically configured hydrophone array designed for towing in combination with real-time PAMGuard software. The real-time detection of porpoises is further indisputable evidence augmenting previous T-POD detections around rigs and platforms [10], proving animals’ presence in the near vicinity of O&G installations’ legs.

Finally, this paper presents the first evidence of porpoises potentially targeting a jack-up rig compared to natural, open sea

habitat (control), supporting the hypothesis that the rig may be acting as an artificial reef, potentially supporting porpoise prey species, and therefore an easily-accessible (and protected from commercial fishing) food resource. Further research has shown that known porpoise prey species were present around some of the installations listed in this study [27]. 5. CONCLUSIONS

Porpoises in various stages of their life cycle are sighted from jack-up rigs and production platforms during routine operations. While the evidence is leaning towards the theory that porpoises targeted a jack-up rig as a potential foraging location in preference to the open sea, more replicated and controlled visual and acoustic data are required to test this hypothesis accurately. If these data are to be presented to the Oslo and Paris Convention (OSPAR) in 2017 as further evidence for a potential rigs-to-reefs programme in the North Sea, then industry needs to fund dedicated scientific studies, as opposed to relying on industrial scientists self-funding in order to turn routine monitoring for permit compliance into research opportunities. Careful, pre-planning, involving participating contracted qualified and experienced scientists needs to be undertaken in order to present further robust data to the peer-reviewed scientific community. Nonetheless, this paper presents further evidence to support the notion that removal of valuable porpoise foraging habitat in the form of offshore O&G installations due to decommissioning at the end of their operational lifetimes may have implications for the long-term survival of this species in the North Sea. 6. ACKNOWLEDGEMENTS

Thanks to our clients for releasing OSC from Non-disclosure Agreement. OSC funded the controlled T-POD experiment, and all data preparation, analysis and presentation for this paper.

7. REFERENCES 1. Todd V.L.G., Todd I.B., Gardiner J.C., Morrin E.C.N.,

2015 'Marine Mammal Observer and Passive Acoustic Monitoring Handbook' Pelagic Publishing Ltd, UK,395

2. Todd V.L.G., White P.R., 2010 'Proximate measurements of acoustic emissions associated with the installation and operation of an exploration jack-up drilling-rig in the North Sea' In: Popper A, Hawkins T, editors. 2nd International Conference on the effects of noise on aquatic marine life on 15-20th August. Cork: Springer.

3. MacLeod C.D., Santos M.B., Reid R.J., Scott B.E., Pierce G.J., 2007 'Linking sandeel consumption and the likelihood of starvation in harbour porpoises in the Scottish North Sea: could climate change mean more starving porpoises?', Proceedings of the Royal Society Biology Letters 3, 185-188

4. Fontaine M.C., Tolley K.A., Michaux J.R., Birkun A., Ferreira M., Jauniaux T., Llavona A., Ozturk B., Ozturk A.A., Ridoux V., Rogan E., Sequeira M., Bouquegneau J.M., Baird S.J.E., 2010 'Genetic and historic evidence for climate-driven population fragmentation in a top cetacean predator: the harbour porpoises in European water', Proceedings of the Royal Society B-Biological Sciences, 277, 2829-2837.

5. Vinther M., 1999 'Bycatches of harbour porpoises, Phocoena phocoena (L.) in Danish set-net fisheries', Journal of Cetacean Research and Management, 1, 123-135.

6. Berggren P., Ishaq R., Zebuhr Y., Naf C., Bandh C., Broman D., 1999 'Patterns and levels of organochlorines (DDTs, PCBs, non-ortho PCBs and PCDD/Fs) in male harbour porpoises (Phocoena phocoena) from the Baltic Sea, the Kattegat-Skagerrak Seas and the west coast of Norway', Marine Pollution Bulletin, 38, 1070-1084.

7. Fabi G., Grati F., Puletti M., Scarcella G., 2004 'Effects on fish community induced by installation of two gas platforms in the Adriatic Sea', Marine Ecology Progress Series, 273, 187-197.

8. Love M.S., Caselle J.E., Snook L., 2000 'Fish assemblages around seven oil platforms in the Santa Barbara channel area', Fishery Bulletin, 98, 96-117.

9. Bell N., Smith J., 1999 'Coral growing on North Sea oil rigs', Nature, 402, 601-601.

10. Todd V.L.G., Pearse W.D., Tregenza N.C., Lepper P.A., Todd I.B., 2009 'Diel echolocation activity of harbour porpoises (Phocoena phocoena) around North Sea offshore gas installations', ICES Journal of Marine Science, 66, 734 - 745.

11. Todd V.L.G., Todd I.B., Gardiner J.C., 2015 'Meals on wheels? A decade of megafaunal visual and

towed Passive Acoustic Monitoring detections from on-tow and stationary offshore drilling rigs and production platforms in the North and Irish Seas', PloS One, Submitted.

12. Jian J., Todd V.L.G., Gardiner J.C., I.B. T., 2015 'Measurements of underwater conductor hammering noise: compliance with the German UBA limit and relevance to the harbour porpoise (Phocoena phocoena)', Proceedings of the 10th European Congress and Exposition on Noise Control Engineering, PACS no. 43.30.+m, 92.20.Jt, 6.

13. JNCC, 2010 'JNCC guidelines for minimising the risk of injury and disturbance to marine mammals from seismic surveys',Aberdeen: Joint Nature Conservation Committee. 16 p.

14. Villadsgaard A., Wahlberg M., Tougaard J., 2007 'Echolocation signals of wild harbour porpoises, Phocoena phocoena', Journal of Experimental Biology, 210, 56-64.

15. Teilmann J., Miller L.A., Kirketerp T., Kastelein R.A., Madsen P.T., Nielsen B.K., Au W.W.L., 2002 'Characteristics of echolocation signals used by a harbour porpoise (Phocoena phocoena) in a target detection experiment', Aquatic Mammals, 28, 275-284.

16. Au W.W.L., 1999 'Transmission beam pattern and echolocation signals of a harbour porpoise (Phocoena phocoena)', Journal of the Acoustical Society of America, 106, 3699-3705.

17. Goodson A.D., Sturtivant C.R., 1996 'Sonar characteristics of the harbour porpoise (Phocoena phocoena): Source levels and spectrum', ICES Journal of Marine Science, 53, 465-472.

18. Hammond P.S., Berggren P., Benke H., Borchers D.L., Collet A., Heide-Jørgensen M.P., Heimlich S., Hiby A.R., Leopold M.F., Øien N., 2002 'Abundance of harbour porpoise and other cetaceans in the North Sea and adjacent waters', Journal of Applied Ecology, 39, 361-376.

19. Palka D., 1996 'Effects of Beaufort sea state on the sightability of harbour porpoises in the Gulf of Maine', Reports of the International Whaling Commission, 46, 575-582.

20. Kastelein R.A., Verlaan M., Jennings N., 2008 'Number and duration of echolocation click trains produced by a harbor porpoise (Phocoena phocoena) in relation to target and performance', Journal of Acoustical Society of America, 124, 40-43.

21. Clausen K.T., Wahlberg M., Beedholm K., DeRuiter S.L., Madsen P.T., 2010 'Click communicatrion in harbour porpoises Phocoena phocoena', Bioacoustics, 20, 1-28.

22. Kastelein R.A., Hardeman J., Boer H., editors (1997) Food consumption and body weight of harbour porpoises (Phocoena phocoena). Woerden The Netherlands: De Spil B.V. Publishers. 217-233. p.

23. Jepson P.U., 2001 'Denmark progress report on cetacean research, April 2001-April 2002', Reports of the International Whaling Commission, IWC/SC/53/.

24. Koopman H.N., 1994 'Topograhical distribution and fatty acid composition of blubber in the harbour porpoise (Phocoena phocoena). MSc Thesis'. Guelph, Ontario: University of Guelph.

25. Otani S., Naito Y., Kato A., Kawamura A., 2000 'Diving behavior and swimming speed of a free-ranging harbor porpoise, Phocoena phocoena', Marine Mammal Science, 16, 811-814.

26. Linnenschmidt M., Teilmann J., Akamatsu T., Dietz R., Miller L.A., 2013 'Biosonar, dive, and foraging activity of satellite tracked harbor porpoises (Phocoena phocoena)', Marine Mammal Science, 29, E77-E97.

27. Lavallin E.W., Todd, V. L. G., Gardiner, J. C. & Hall, S., 2015 ' A quantitative analysis of invertebrate and fish assemblage dynamics in association with a North Sea oil & gas installation complex using ROV visual inspection data.', ICES Journal of Marine Science Submitted.